-
Notifications
You must be signed in to change notification settings - Fork 0
/
run_lora_exp_train.sh
106 lines (101 loc) · 3.85 KB
/
run_lora_exp_train.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
data_dir=./sample_data
nmd=/vol/bitbucket/l22/QKV_llama2_models_lora_16
pof=./QKV_predictions_llama2_lora_16.txt
log=./log_files/QKV_log_llama2_lora_exp.txt
ckpt_path=/vol/bitbucket/l22/QKV_llama2_models_lora_16/ckpt_0
accelerate launch ./src/relation_extraction.py \
--model_type llama2 \
--data_format_mode 0 \
--classification_scheme 2 \
--pretrained_model meta-llama/Llama-2-7b-hf \
--data_dir $data_dir \
--num_core 3 \
--new_model_dir $nmd \
--predict_output_file $pof \
--overwrite_model_dir \
--seed 13 \
--max_seq_length 256 \
--cache_data \
--do_train \
--do_lower_case \
--lora_rank 16 \
--lora_alpha 32 \
--train_batch_size 2 \
--eval_batch_size 2 \
--learning_rate 1e-4 \
--num_train_epochs 3 \
--gradient_accumulation_steps 4 \
--do_warmup \
--warmup_ratio 0.1 \
--weight_decay 0 \
--max_num_checkpoints 2 \
--log_file $log \
--log_step 500 \
--progress_bar
data_dir=./sample_data
nmd=/vol/bitbucket/l22/QKV_llama2_models_lora_8
pof=./QKV_predictions_llama2_lora_8.txt
ckpt_path=/vol/bitbucket/l22/QKV_llama2_models_lora_8/ckpt_0
accelerate launch ./src/relation_extraction.py \
--model_type llama2 \
--data_format_mode 0 \
--classification_scheme 2 \
--pretrained_model meta-llama/Llama-2-7b-hf \
--data_dir $data_dir \
--num_core 3 \
--new_model_dir $nmd \
--predict_output_file $pof \
--overwrite_model_dir \
--seed 13 \
--max_seq_length 256 \
--cache_data \
--do_train \
--do_lower_case \
--lora_rank 8 \
--lora_alpha 32 \
--train_batch_size 2 \
--eval_batch_size 2 \
--learning_rate 1e-4 \
--num_train_epochs 3 \
--gradient_accumulation_steps 4 \
--do_warmup \
--warmup_ratio 0.1 \
--weight_decay 0 \
--max_num_checkpoints 2 \
--log_file $log \
--log_step 500 \
--progress_bar
data_dir=./sample_data
nmd=/vol/bitbucket/l22/QKV_llama2_models_lora_4
pof=./QKV_predictions_llama2_lora_4.txt
ckpt_path=/vol/bitbucket/l22/QKV_llama2_models_lora_4/ckpt_0
# NOTE: we have more options available, you can check our wiki for more information
accelerate launch ./src/relation_extraction.py \
--model_type llama2 \
--data_format_mode 0 \
--classification_scheme 2 \
--pretrained_model meta-llama/Llama-2-7b-hf \
--data_dir $data_dir \
--num_core 3 \
--new_model_dir $nmd \
--predict_output_file $pof \
--overwrite_model_dir \
--seed 13 \
--max_seq_length 256 \
--cache_data \
--do_train \
--do_lower_case \
--lora_rank 4 \
--lora_alpha 32 \
--train_batch_size 2 \
--eval_batch_size 2 \
--learning_rate 1e-4 \
--num_train_epochs 3 \
--gradient_accumulation_steps 4 \
--do_warmup \
--warmup_ratio 0.1 \
--weight_decay 0 \
--max_num_checkpoints 2 \
--log_file $log \
--log_step 500 \
--progress_bar